https://aiweekly.co/alerts/3m-expert-used-chatgpt-to-draft-0-at-fault-defense-report AI Weekly en 2026-08-17T16:57:16+00:00 3M expert used ChatGPT to draft '0% at fault' defense report https://aiweekly.co/alerts/svopd-lifts-qwen35-4b-above-qwen3-vl-235b-and-gpt-54 AI Weekly en 2026-08-17T15:54:56+00:00 S²VOPD lifts Qwen3.5-4B above Qwen3-VL 235B and GPT-5.4 https://aiweekly.co/alerts/llms-mirror-the-brains-modular-neuron-layout-study-finds AI Weekly en 2026-08-17T15:54:56+00:00 LLMs Mirror the Brain's Modular Neuron Layout, Study Finds https://aiweekly.co/alerts/paper-flags-amplification-lift-gap-in-reasoning-trained-models AI Weekly en 2026-08-17T15:54:56+00:00 Paper flags 'Amplification-Lift Gap' in reasoning-trained models https://aiweekly.co/alerts/apodex-discovery-beats-sota-on-aav-capsid-design-by-7 AI Weekly en 2026-08-17T15:54:56+00:00 Apodex Discovery beats SOTA on AAV capsid design by 7% https://aiweekly.co/alerts/paper-diagnoses-forecast-collapse-in-tsfm-stock-ranking AI Weekly en 2026-08-17T15:54:56+00:00 Paper Diagnoses 'Forecast Collapse' in TSFM Stock Ranking https://aiweekly.co/alerts/github-outage-hits-web-api-actions-and-copilot-worldwide AI Weekly en 2026-08-17T15:29:18+00:00 GitHub outage hits web, API, Actions and Copilot worldwide https://aiweekly.co/alerts/gravis-robotics-raises-200m-from-softbank-to-retrofit-excavators AI Weekly en 2026-08-17T15:29:18+00:00 Gravis Robotics raises $200M from SoftBank to retrofit excavators https://aiweekly.co/alerts/amazon-scans-and-destroys-rare-books-at-vegas-ai-facility AI Weekly en 2026-08-17T15:12:01+00:00 Amazon Scans and Destroys Rare Books at Vegas AI Facility https://aiweekly.co/alerts/littlelearner-a-5b-llm-bounded-to-us-grade-5-material AI Weekly en 2026-08-17T15:12:01+00:00 LittleLearner: a 5B LLM bounded to U.S. Grade-5 material https://aiweekly.co/alerts/trumps-world-liberty-backs-worldclaw-chinese-ai-gateway AI Weekly en 2026-08-17T13:58:30+00:00 Trump's World Liberty backs WorldClaw, Chinese AI gateway https://aiweekly.co/alerts/groq-raises-350m-at-35b-half-prior-value-after-nvidia-deal AI Weekly en 2026-08-17T13:58:30+00:00 Groq Raises $350M at $3.5B, Half Prior Value After Nvidia Deal https://aiweekly.co/alerts/chip-giants-back-250b-of-ai-rounds-in-2026-openai-95 AI Weekly en 2026-08-17T13:58:30+00:00 Chip Giants Back $250B of AI Rounds in 2026, OpenAI 95% https://aiweekly.co/alerts/openai-models-escape-sandbox-hack-hugging-face-for-benchmark AI Weekly en 2026-08-17T12:36:17+00:00 OpenAI models escape sandbox, hack Hugging Face for benchmark https://aiweekly.co/alerts/wispr-closes-280m-series-b-at-2b-valuation-led-by-menlo AI Weekly en 2026-08-17T12:29:01+00:00 Wispr Closes $280M Series B at $2B Valuation Led by Menlo https://aiweekly.co/alerts/meta-blackrock-14b-el-paso-data-center-only-partially-insured AI Weekly en 2026-08-17T11:00:17+00:00 Meta-BlackRock $14B El Paso data center only partially insured https://aiweekly.co/alerts/pjms-4-gw-modeling-gap-cost-ratepayers-12b-semianalysis-says AI Weekly en 2026-08-17T09:28:40+00:00 PJM's 4 GW modeling gap cost ratepayers ~$12B, SemiAnalysis says https://aiweekly.co/alerts/aronofsky-blomkamp-build-ai-studios-on-us-chinese-models AI Weekly en 2026-08-17T09:28:40+00:00 Aronofsky, Blomkamp Build AI Studios on US, Chinese Models https://aiweekly.co/alerts/khosla-backs-discovery-loops-bet-on-ai-as-the-researcher AI Weekly en 2026-08-17T09:12:01+00:00 Khosla backs Discovery Loop's bet on AI as the researcher https://aiweekly.co/alerts/disney-netflix-and-amazon-ramp-ai-production-hiring AI Weekly en 2026-08-17T06:12:01+00:00 Disney, Netflix and Amazon Ramp AI Production Hiring https://aiweekly.co/alerts/ra-bench-19-detectors-fail-on-ai-generated-crisis-videos AI Weekly en 2026-08-17T04:03:55+00:00 RA-Bench: 19 Detectors Fail on AI-Generated Crisis Videos https://aiweekly.co/alerts/paper-frontier-agents-adapt-rd-methods-rarely-invent-new-ones AI Weekly en 2026-08-17T04:03:55+00:00 Paper: Frontier Agents Adapt R&D Methods, Rarely Invent New Ones https://aiweekly.co/alerts/pnas-peer-take-up-lifts-parent-ai-spend-60-warnings-dont AI Weekly en 2026-08-17T04:03:55+00:00 PNAS: Peer Take-Up Lifts Parent AI Spend 60%, Warnings Don't https://aiweekly.co/alerts/singapore-bets-ai-access-keeps-fund-managers-from-hong-kong AI Weekly en 2026-08-17T03:28:56+00:00 Singapore Bets AI Access Keeps Fund Managers From Hong Kong https://aiweekly.co/alerts/qwen-38-27b-strong-open-model-wildly-overthinks-by-default AI Weekly en 2026-08-17T03:12:01+00:00 Qwen 3.8 27B: strong open model, wildly overthinks by default https://aiweekly.co/alerts/anthropic-confirms-42-minute-outage-on-claudes-web-apps AI Weekly en 2026-08-17T02:00:28+00:00 Anthropic Confirms 42-Minute Outage on Claude's Web Apps https://aiweekly.co/alerts/cohere-labs-ships-24b-open-weight-vision-language-model AI Weekly en 2026-08-17T00:12:02+00:00 Cohere Labs ships 2.4B open-weight vision-language model https://aiweekly.co/alerts/brennan-center-chatbots-reject-fraud-myths-botch-citations AI Weekly en 2026-08-17T00:12:02+00:00 Brennan Center: Chatbots Reject Fraud Myths, Botch Citations https://aiweekly.co/alerts/deepseek-v4-pro-exits-preview-with-mit-licensed-17t-weights AI Weekly en 2026-08-17T00:12:02+00:00 DeepSeek V4 Pro exits preview with MIT-licensed 1.7T weights https://aiweekly.co/alerts/white-house-to-extend-ai-safety-framework-to-open-models AI Weekly en 2026-08-17T00:12:02+00:00 White House to Extend AI Safety Framework to Open Models https://aiweekly.co/alerts/cohere-labs-debuts-north-micro-vision-24b-native-res-vlm AI Weekly en 2026-08-17T00:12:02+00:00 Cohere Labs debuts North Micro Vision, 2.4B native-res VLM https://aiweekly.co/alerts/googles-gemini-37-flash-targets-coding-and-agent-work AI Weekly en 2026-08-17T00:12:02+00:00 Google's Gemini 3.7 Flash targets coding and agent work https://aiweekly.co/alerts/meta-superintelligence-lab-open-sources-muse-glimmer-30b AI Weekly en 2026-08-17T00:12:02+00:00 Meta Superintelligence Lab open-sources Muse Glimmer-30B https://aiweekly.co/alerts/minimax-open-sources-music-3-a-qwen3-based-song-generator AI Weekly en 2026-08-17T00:12:02+00:00 MiniMax open-sources Music 3, a Qwen3-based song generator https://aiweekly.co/alerts/predicted-ai-job-apocalypse-hasnt-arrived-guardian-finds AI Weekly en 2026-08-17T00:12:02+00:00 Predicted AI job apocalypse hasn't arrived, Guardian finds https://aiweekly.co/alerts/trace-inversion-extracts-hidden-reasoning-from-gpt-5-mini AI Weekly en 2026-08-17T00:12:02+00:00 Trace Inversion Extracts Hidden Reasoning From GPT-5 mini https://aiweekly.co/alerts/full-bandwidth-transformer-claims-15x-token-efficiency AI Weekly en 2026-08-17T00:12:02+00:00 'Full-bandwidth transformer' claims 1.5x token efficiency https://aiweekly.co/alerts/model-discovery-agent-pairs-llm-proposals-with-bayesian-tests AI Weekly en 2026-08-17T00:12:02+00:00 Model Discovery Agent pairs LLM proposals with Bayesian tests https://aiweekly.co/alerts/xiaohongshu-opens-dots3-note-a-280b-moe-multimodal-model AI Weekly en 2026-08-17T00:12:02+00:00 Xiaohongshu opens dots3-note, a 280B MoE multimodal model https://aiweekly.co/alerts/bcsd-trains-llm-agents-to-actually-consult-their-skill-libraries AI Weekly en 2026-08-17T00:12:02+00:00 BCSD trains LLM agents to actually consult their skill libraries https://aiweekly.co/alerts/matryoshka-llm-suite-bundles-500m-3b-into-one-nested-stack AI Weekly en 2026-08-17T00:12:02+00:00 Matryoshka LLM suite bundles 500M-3B into one nested stack https://aiweekly.co/alerts/skillzip-shrinks-agent-skill-libraries-without-evaluation-runs AI Weekly en 2026-08-17T00:12:02+00:00 SkillZip shrinks agent skill libraries without evaluation runs https://aiweekly.co/alerts/smart-birhane-audit-finds-ai-alignment-rarely-defines-values AI Weekly en 2026-08-17T00:12:02+00:00 Smart, Birhane audit finds AI alignment rarely defines values https://aiweekly.co/alerts/tsugo-uses-go-life-and-death-puzzles-to-grade-llm-search AI Weekly en 2026-08-17T00:12:02+00:00 TsuGO uses Go life-and-death puzzles to grade LLM search https://aiweekly.co/alerts/antirez-ships-465-gb-q2-deepseek-v4-pro-gguf-for-ds4-engine AI Weekly en 2026-08-17T00:12:02+00:00 antirez ships 465 GB Q2 DeepSeek V4 Pro GGUF for DS4 engine https://aiweekly.co/alerts/cnbc-poll-81-of-young-americans-distrust-palantirs-karp-on-ai AI Weekly en 2026-08-16T22:58:55+00:00 CNBC poll: 81% of young Americans distrust Palantir's Karp on AI https://aiweekly.co/alerts/stripe-finalizes-reported-7b-buy-of-ai-gateway-openrouter AI Weekly en 2026-08-16T21:32:23+00:00 Stripe finalizes reported $7B+ buy of AI gateway OpenRouter https://aiweekly.co/alerts/meta-glasses-harassment-videos-still-on-instagram-after-ban AI Weekly en 2026-08-16T18:12:01+00:00 Meta glasses harassment videos still on Instagram after ban https://aiweekly.co/alerts/inaudible-frequency-attack-cuts-audio-llm-accuracy-up-to-67pp AI Weekly en 2026-08-16T15:53:17+00:00 Inaudible-frequency attack cuts audio-LLM accuracy up to 67pp https://aiweekly.co/alerts/gambit-beam-search-cuts-reasoning-tokens-685-adds-67pp AI Weekly en 2026-08-16T15:53:17+00:00 Gambit Beam Search Cuts Reasoning Tokens 68.5%, Adds 6.7pp https://aiweekly.co/alerts/uniswap-unifies-face-and-voice-swap-in-one-diffusion-model AI Weekly en 2026-08-16T15:53:17+00:00 UniSwap unifies face and voice swap in one diffusion model https://aiweekly.co/alerts/techcrunch-maps-session-checks-for-chatgpt-claude-perplexity AI Weekly en 2026-08-16T14:00:00+00:00 TechCrunch Maps Session Checks for ChatGPT, Claude, Perplexity https://aiweekly.co/alerts/gamgee-launches-at-yc-s26-with-ai-designed-dog-cancer-vaccines AI Weekly en 2026-08-16T09:31:51+00:00 Gamgee launches at YC S26 with AI-designed dog cancer vaccines https://aiweekly.co/alerts/nature-paper-argues-open-ai-still-runs-on-closed-stacks AI Weekly en 2026-08-16T09:12:01+00:00 Nature paper argues 'open' AI still runs on closed stacks https://aiweekly.co/alerts/darwinx-evolves-agent-harnesses-to-93-on-webarena-model-frozen AI Weekly en 2026-08-16T03:54:45+00:00 DarwinX evolves agent harnesses to 93% on WebArena, model frozen https://aiweekly.co/alerts/dreamx-phi-10-wins-track-1-of-the-worldarena-20-challenge AI Weekly en 2026-08-16T03:54:45+00:00 DreamX-Phi 1.0 Wins Track 1 of the WorldArena 2.0 Challenge https://aiweekly.co/alerts/bond-traders-fret-over-70b-in-ai-shadow-credit-backstops AI Weekly en 2026-08-16T03:28:47+00:00 Bond Traders Fret Over $70B in AI Shadow Credit Backstops https://aiweekly.co/alerts/nyc-ban-the-scan-bill-wins-mamdani-council-majority AI Weekly en 2026-08-16T03:12:01+00:00 NYC 'Ban the Scan' Bill Wins Mamdani, Council Majority https://aiweekly.co/alerts/raji-and-bender-challenge-the-general-ai-benchmark-myth AI Weekly en 2026-08-16T03:12:01+00:00 Raji and Bender challenge the 'general' AI benchmark myth https://aiweekly.co/alerts/georgia-techs-topaz-adds-audit-trails-to-agentic-model-routing AI Weekly en 2026-08-16T00:12:01+00:00 Georgia Tech's Topaz Adds Audit Trails to Agentic Model Routing https://aiweekly.co/alerts/nvidia-discloses-21b-spacex-stake-30b-intel-position AI Weekly en 2026-08-15T23:00:51+00:00 Nvidia Discloses $21B SpaceX Stake, $30B Intel Position https://aiweekly.co/alerts/mass-teen-used-chatgpt-for-murder-fantasy-stories-da-says AI Weekly en 2026-08-15T23:00:51+00:00 Mass. teen used ChatGPT for murder 'fantasy stories,' DA says https://aiweekly.co/alerts/anthropic-details-claude-watermarks-gaps-on-code-and-rewrites AI Weekly en 2026-08-15T23:00:51+00:00 Anthropic Details Claude Watermark's Gaps on Code and Rewrites https://aiweekly.co/alerts/brazils-anpd-halts-facial-recognition-in-paran-schools AI Weekly en 2026-08-15T21:12:01+00:00 Brazil's ANPD Halts Facial Recognition in Paraná Schools https://aiweekly.co/alerts/claude-max-subscribers-cancel-over-anthropics-new-watermark AI Weekly en 2026-08-15T21:12:01+00:00 Claude Max subscribers cancel over Anthropic's new watermark https://aiweekly.co/alerts/nvidia-weighs-up-to-3b-stake-in-softbank-power-arm-sb-energy AI Weekly en 2026-08-15T19:59:35+00:00 Nvidia weighs up to $3B stake in SoftBank power arm SB Energy https://aiweekly.co/alerts/mercor-chases-shuttered-startups-slack-and-jira-archives AI Weekly en 2026-08-15T19:59:35+00:00 Mercor chases shuttered startups' Slack and Jira archives https://aiweekly.co/issues/ai-ethics-is-nobodys-job-now-the-labs-prefer-it-that-way AI Weekly en 2026-08-17T12:00:00+00:00 AI ethics is nobody's job now. The labs prefer it that way. https://aiweekly.co/issues/zuckerberg-promises-superintelligence-for-all-experts-arent AI Weekly en 2026-08-16T12:00:00+00:00 Zuckerberg promises superintelligence for all. Experts aren't sold.